Coordinated Exploration in Distributed Reinforcement Learning
Keywords: Coordinated Exploration, Distributed Reinforcement Learning
Abstract: We propose Collective Optimism, a new method for Coordinated Exploration in Distributed Reinforcement Learning. The primary objective of collective optimism is to efficiently coordinate agents' exploration strategies, a concept that has not been previously considered in existing models. This algorithm ensures the maximization of overall performance by taking into account the uncertainty of the current state-action pair, without relying on randomness. A strength of this method is its remarkable ability to be highly applicable to any RL model. To demonstrate the validity of our method and to show its adoptability, we implemented and evaluated the performance of CODRL model through a series of numerical experiments.
Primary Area: reinforcement learning
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.
Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.
Reciprocal Reviewing: I understand the reciprocal reviewing requirement as described on https://iclr.cc/Conferences/2025/CallForPapers. If none of the authors are registered as a reviewer, it may result in a desk rejection at the discretion of the program chairs. To request an exception, please complete this form at https://forms.gle/Huojr6VjkFxiQsUp6.
Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.
No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.
Resubmission: Yes
Student Author: Yes
Large Language Models: Yes, at the sentence level (e.g., fixing grammar, re-wording sentences)
Submission Number: 13464
Loading